Expanding Network Analysis Tools in Psychological Networks: Minimal Spanning Trees, Participation Coefficients, and Motif Analysis Applied to a Network of 26 Psychological Attributes

Abstract

The analysis of psychological networks in previous research has been limited to the inspection of centrality measures and the quantification of specific global network features. The main idea of this paper is that a psychological network entails more potentially useful and interesting information that can be reaped by other methods widely used in network science. Specifically, we suggest methods that provide clearer picture about hierarchical arrangement of nodes in the network, address heterogeneity of nodes in the network, and look more closely at network’s local structure. We explore the potential value of minimum spanning trees, participation coefficients, and motif analyses and demonstrate the relevant analyses using a network of 26 psychological attributes. Using these techniques, we investigate how the network of different psychological concepts is organized, which attribute is most central, and what the role of intelligence in the network is relative to other psychological variables. Applying the three methods, we arrive at several tentative conclusions. Trait Empathy is the most “central” attribute in the network. Intelligence, although peripheral, is weakly but equally related to different kinds of attributes present in the network. Analysis of triadic configurations additionally shows that the network is characterized by relatively strong open triads and an unusually frequent occurrence of negative triangles. We discuss these and other findings in the light of possible theoretical explanations, methodological limitations, and future research.

Network structure of depression symptomology in participants with and without depressive disorder: the population-based Health 2000–2011 study

“…There are multiple problems in interpreting central symptoms as the most influential, (the most central may be just the end point or just the one with the greatest variability, see: https://psych-networks.com/how-to-not-interpret-centrality-values-in-network-structures/ ) and recently the whole basis of measuring centrality in psychological networks that do not have similar features (serial flow of connections) as social networks, has been challenged [ 20 ]. In the present study, we tried to overcome some of these problems using centrality measures that are not based on shortest path measures (strength centrality) and by taking into account the community structure within the network (participation coefficient) [ 41 ].…”

Network structure of psychotic‐like experiences in adolescents: Links with risk and protective factors

“…Finally, network modelling is currently in its initial stages (Guloksuz, Pries, & van Os, 2017). Although it is shown as a promising tool in obtaining information in a number of research fields, there remain to be limitations and resolution issues (Jones, Heeren, & McNally, 2017;Letina, Blanken, Deserno, & Borsboom, 2019).…”

Does mindfulness change the mind? A novel psychonectome perspective based on Network Analysis

“…Beyond these specific results, the current ‘psychonectome’ proof-of-concept approach seems useful to provide further evidence of the mind as a complex network of psychological constructs. The approach of considering topological aspects of the relations between constructs may help to enhance our understanding of psychological functioning as a complex network of interacting elements that are mutually interconnected [18–20,137,138]. This unique perspective has become central in fields like neuroscience [25,35] and is likely that can also provide useful insights on the functioning and mechanisms of human mind.…”

“…Nevertheless, some network theorists have recently suggested that dwelling excessively on the lack of interest of latent variables may misguide the focus from what is more promising in conceptualizing disorders from a NA perspective (i.e., the idea of causally interconnected elements) [16,17]. In this sense, the use of psychological constructs (e.g., traits assessed by questionnaires), and not only elements like symptoms or signs, is beginning to be used in NA (e.g., [18–20]) and it could also be possible that, in the future, hybrid models, using not only elements like signs and symptoms but also latent variables, could shed light on the connections between psychological elements [21,22].…”